PhD student in Remote Sensing / Deep Learning
Munich
Joined April 2010
- Tweets65
- Following154
- Followers117
- Likes619
Konrad Heidler retweeted
This paper has been a long time in the making. It represents an important step towards the development of intelligent processing strategies for #glacier monitoring in #Greenland.
doi.org/10.5194/tc-18-3315-2…
It's been an amazing journey, and I am truly grateful for all the experiences made and people met along the way. Very excited for the next adventures to come! 🚀
For an overview of what I've been doing during my PhD, check out the project pages here:
konrad.heidler.info/disserta…
Very proud to announce that I am now the owner of a fancy new hat! #PhDone 🎓
Thanks to my great committee, Profs. @xiaoxiang_zhu, Richard Bamler, Sébastien Lefèvre, and @jlbamber for your time, commitment and support!
Konrad Heidler retweeted
🚨New paper in @ESSD_journal ! Using deep learning, we created a new calving front dataset for Svalbard with an unprecedented spatiotemporal resolution. It includes almost 125000 terminus traces from 149 marine-terminating glaciers, covering the years 1985 to 2023. (1/4)
Just published, superb, comprehensive calving data set for Svalbard, all Open Access of course, @icetianli , @xiaoxiang_zhu
essd.copernicus.org/articles…
Konrad Heidler retweeted
Just published, superb, comprehensive calving data set for Svalbard, all Open Access of course, @icetianli , @xiaoxiang_zhu
essd.copernicus.org/articles…
Konrad Heidler retweeted
Our latest paper about out today in @Nature is all about calving-front retreat in Greenland.
Research Briefing: rdcu.be/dwffe
Full article: rdcu.be/dwfcq
📚 Just published! ieeexplore.ieee.org/document…
Our new paper rethinks glacier calving front delineation with neural networks. With COBRA, we say goodbye to pixel-wise predictions! 🏔️🐍
How do we do that? [1/6]
@MouLichao, @ErikLoebel, M. Scheinert, S. Lefèvre, @xiaoxiang_zhu
📈📉 As an added bonus, the explicit contour detection approach allows for better uncertainty quantification than pixel-wise approaches.
Some dozens of vertices are a much cleaner representation of a calving front than a 2d array of pixel-wise prediction probabilities. [5/6]
🗺️ Want to see some COBRA predictions in the wild? Check out our inference map: khdlr.github.io/COBRA/map.ht… [6/6]
Konrad Heidler retweeted
📰 "Growing permafrost slump threatens part of Alaska Highway, road to be moved" @CBCNews @CBCNorth
@cpa_acp @YukonUniversity @yukongov @YukonHPW
cbc.ca/news/canada/north/per…
Konrad Heidler retweeted
Maximum size of a PDF, version 7: 381 km × 381 km.
commons.m.wikimedia.org/wiki…
Konrad Heidler retweeted
🚨 Preprint alert 🚨
We use Universal Differential Equations, embedding neural networks in differential equations describing glacier ice flow, to learn hidden physical laws from data. First case study of @ODINN_org, as a new @JuliaLanguage package.
👇
gmd.copernicus.org/preprints…
Konrad Heidler retweeted
An exciting first application of our #DeepLearning based #Greenland #Glacier Front product! Take a look at how the calving style of Ingia Isbræ changes as it retreats into the topographic minimum. Reduction in both seasonal amplitude and retreat rate, but for how long?
Konrad Heidler retweeted
This paper synthesizes the results of my research over the past three years. It has been a long time in the making, and I am excited that it is finally submitted and open for discussion @EGU_TC.
Discussion open for comments: Calving front monitoring at sub-seasonal resolution: a deep learning application to Greenland glaciers doi.org/10.5194/tc-2023-52
Konrad Heidler retweeted
My first experimental #streamlit #app for plotting #weather/#climate series based on my noaaplotter #python package
#dataviz #geo #matplotlib
Link: initze-noaaplotter-streamlit…
Repos:
github.com/initze/noaaplotte…
github.com/initze/noaaplotte…
